Developer Softacom has released Delphi YOLO ONNX Runtime Wrapper, a library that brings YOLO-based object detection and computer vision capabilities directly to Delphi applications using ONNX Runtime. The framework allows developers to run trained AI models and detect objects in images without leaving the Object Pascal ecosystem.
The library provides a native Delphi interface for loading ONNX models, running inference, and processing detection results, making it easier to integrate modern AI-powered vision features into both VCL and FireMonkey applications.
The framework includes support for:
- YOLO object detection models
- ONNX Runtime integration
- Image classification and recognition
- Model metadata extraction
- CPU and GPU inference
- VCL and FireMonkey applications
- Windows, Linux, and macOS deployments
The project uses ONNX Runtime and supports multiple execution providers, including DirectML, CUDA, and CPU inference, allowing developers to choose the best runtime for their deployment environment.
Whether you’re building video analytics, smart cameras, industrial automation systems, or AI-powered desktop applications, Delphi YOLO ONNX Runtime Wrapper provides a practical way to bring computer vision workflows to Delphi.
See how to run YOLO object detection models from Delphi using ONNX Runtime.
